ARGON-ETHYLENE LAYERFORMATION ON GRAPHITIZED CARBONBLACK
969
Argon-Ethylene Layer Formation on Graphitized Carbon Blackla by Carl F. Prenzlow,lb H. Richard Beard, and Robert S. Brundage National Bureau of Standards, Gaithersberg, Maryland, and California State College at Fullerton, Fullerton, California (Received December 18, 1967)
Argon isotherms have been measured on various coverages of ethylene on graphitized carbon black in the range 6678°K. The molecular area of ethylene at 77°K was estimated to be 17.8-18.8 bz. This value corresponds best with a model of ethylene molecules rocking to and fro on the carbon surface. The prefrozen layer methods of Halsey have been extended to a system in which capillarity is a complication. The maximum thickness of the ethylene film was found to be 2 layers. Argon V , lowering on ethylene was observed and attributed to loose packing of argon rather than to ethylene capillarity or surface roughness of carbon. A two-dimensional phase transition, presumably of the liquid + solid type, has been observed for the first time on a prefrozen layer system.
Introduction The preadsorbed layer technique has been introduced into physical adsorption studies by Halsey and Singleton,2 who studied argon adsorption on xenon, and by Steele and Aston,a who studied helium adsorption on argon. To measure argon isotherms on xenon films with varying characteristics, Halsey deposited xenon by a process of slow cooling onto three widely different underlying surfaces : anatase, silver iodide, and graphitized carbon black. His results show that in general prefrozen layers do not grow indefinitely thick, but rather form layers of definite thickness, and that the maximum finite thickness of a prefrozen film of a particular adsorbate (in this case xenon) varies from adsorbent to adsorbent. The argon-xenon-graphitized carbon system is of particular interest because graphitized carbon is believed to present a uniform potential to both xenon and argon. Studies of this system showed that a single prefrozen layer of xenon is quite uniform, that a film composed of two or more layers is somewhat jumbled, and that the maximum thickness of the xenon film is about six layers. For comparison with the xenon system, we have undertaken similar studies of argon adsorption on ethylene layers on graphitized carbon. Ethylene was chosen because of its close similarity in vapor pressure data with xenon and because of its molecular simplicity.
son scientific grade, which was pumped upon at liquid nitrogen temperatures in successive freeze-thaw cycles until it showed a negligible vapor pressure. A mass spectrometric analysis revealed the gas to be 99.5% pure with 0.33% C4 and Cgunsaturated hydrocarbons, 0.09% Cq and Cs saturated hydrocarbons, and 0.09% Cz and C4 alcohols. Adsorbent. A sample of graphitized carbon black designated Sterling FT (2700) was kindly supplied to us by Dr. W. R. Smith of Godfrey L. Cabot, Inc., Cambridge, Mass. This sample is stated to be equivalent to P-33 (2700), a graphitized carbon which was used by Halsey in previous experiments on xenon 1aye1-s.~ To check the identity of these two carbons, an argon isotherm was measured on bare surface FT (2700) a t 73.9"K and compared with a previously measured isotherm on bare surface P-33 (2700) at 73.5"K. The two isotherms were virtually identical. Furthermore, an argon isotherm a t 77.1"K on one xenon layer on Sterling FT (2700) mas identical with an argon isotherm on one xenon layer on P-33 (2700) measured by Prenzlow and HalseyS4 Samples of Sterling FT (2700) were packed in the adsorption tube 'GO a density of 0.59 g/cc. The dead space for each sample at a particular ethylene coverage was measured with reagent helium a t each liquid nitrogen temperature used. A Styrofoam float gauge was used to measure the level of liquid nitrogen in the cyrostat so that dead spaces could be corrected for bath level variations. Outgassing Procedure. Sterling FT (2700) was out-
Experimental Section V a c u u m System. Adsorption measurements were made in a conventional BET constant-volume apparatus. Gas pressures were read with a cathetometer graduated to 0.05 mm and measured in mercury manometers 14 mm in diameter. Gases. Helium, argon, and xenon gases of better than 99.9% purity were obtained in sealed glass bulbs from Air Reductions Sales Co. Ethylene was Mathe-
(1) (a) Presented a t the 146th National Meeting of the American Chemical Society, Denver, Colo., January 1964; (b) to whom correspondence should be addressed at California State College a t Fullerton, Chemistry Department, Fullerton, Calif. (2) J. H. Singleton and G. D. Halsey, Jr., J . Phys. Chem., 58, 330 (1954). (3) (a) W. A. Steele and J. G. Aston, J . Chem. Phys., 23, 1547 (1955): (b) W. A. Steele and J. G. Aston, J . Amer. Chem. SOC.,79, 2393 (1957). (4) C. F. Prenzlow and G. D. Halsey, Jr., J . Phys. Chem., 61, 1158 (1957). Volume 7% Number 4 April 1969
970 gassed for 12 hr at 400” after evacuation to mm or less prior to each set of adsorption measurements. Argon could be removed from surfaces fractionally covered with ethylene by evacuating for 12 hr at liquid nitrogen temperatures, and it could be removed from one or more layers of ethylene by evacuating for 10 min at liquid nitrogen temperatures. No change was detected in the argon isotherms on ethylene systems after outgassing. Measurement of Temperature. Most of the temperatures in the adsorption vessel were measured with an argon vapor pressure thermometer using the data for solid argon of Clark, Din, and Robb.6 A few temperatures for isotherms measured at 52-54 em, which are just above the argon triple point of 83.78OK, were obtained from liquid vapor pressure data of Clark, Din, and RobbS6 Temperatures of nitrogen isotherms were measured using a nitrogen vapor pressure thermometer and the liquid vapor pressure data of Giauque and Clayton.s Temperatures of ethylene isotherms were measured from ethylene vapor pressures using the liquid vapor pressure data of Egan and Kemp.’ Tempera,tures were controlled and measured to within 0.02”K by pumping the liquid nitrogen cryostat in a manner previously d e s ~ r i b e d . ~ Deposition of Ethylene Layers. The adsorption vessel and its silvered dewar for layer deposition have been previously describedm2Measured gas volumes of ethylene (STP) were prefroaen onto the carbon by controlled cooling2 using a methane vapor thermometer to follow the process. Cooling periods of at least 4 hr were allowed for ethylene depositions up to one layer from a temperature of 106°K to 85°K. At least 16 hr was allowed for deposition of ethylene doses of two or more layers. Several millimeters of reagent helium was added to the ethylene to ensure heat transfer during all depositions.
Results The monolayer volume of ethylene on one gram of Sterling FT (2700) was determined by the surface titration method of Singleton and Halsey.2 Increasing doses of ethylene were prefroaen onto the carbon surface, the argon isotherm at each ethylene dosage was measured, and the argon point B was determined for each of these isotherms. The monolayer volume of ethylene was taken to be 2.44 i 0.01 cc/g since a t this coverage the first hump in an argon isotherm (and the point B) disappears, as shown by the 1.00 ethylene layer isotherm in Figure 1. This precise V, value could be used to indicate the packing and orientation of ethylene molecules on the carbon black surface provided that the true specific area of Sterling FT (2700) were accurately known. Such a specific area can be calculated from reliable V, values for argon, krypton, and xenon on P-33 (2700), which were obtained by ‘(step height” and “surface titration” methods by Halsey The Journal of Physical Chemicrtry
C. F. PRENZLOW, H. R. BEARD,AND R. S. BRUNDAGE and c o - ~ o r k e r s . ~Rare ~ ~ gas data seem ideal for calculating the true specific area of Sterling FT (2700) becase the carbon surface is believed to present no potential energy barriers to prevent hexagonal close packing of rare gas atoms* and because simple spherical particles can present no surface orientation problems. The main problem which arises is in determining the correct area occupied per rare gas atom on the carbon surfaces4 This is usually taken to be the area per atom in the hexagonally packed (111) plane of the solid rare gas crystal. I n this case the specific area of carbon may be calculated from rare gas bulk solid density data by means of the formula
S = 0.4110(M/j1)~’~V,,,
(1)
in which Sis the area in square meters per gram of carbon black, M is the gram atomic weight of the rare gas, and p is the density of the rare gas solid at the tempem ture of V , measurement. Density data for solid argon, krypton, and xenon, which are suitable for making these calculations, have recently been published by Eatwell and Smith.g Estimates of the specific area, calculated in this fashion, have been listed in Table I. Table I: Estimated Specific Areas of P-33 (2700) from Rare Gas Density Data“ Adsorbate Argon Krypton Xenon
Temp,
Solid
OK
density
V,
area, m*/g
64.5 78 78
1.679 2.932 3.647
3.66 3.15 2.76
12.45 12.11 12.30
Calcd
a The true specific area, averaged from these estimates, is taken to be 12.3 f 0.2 m2/g.
On the other hand, Lander and recently have published low-energy election diffraction (LEED) data for xenon adsorbed on the basal plane of graphite. Their data indicate that xenon is hexagonally close packed and has a surprisingly low area per atom of 15.7 Az. Using this value and the xenon V m value 2.76 cc/g,4 one estimates the specific area of Sterling FT (2700) to be 11.6 m2/g. By its very nature, LEED should provide a more direct and exact measurement of atomic areas than density measurements and hence a better estimate of the absolute carbon black surface area. However, because LEED data for argon and krypton on the basal (6) Clark, Din, and Robb; Michels, Wasseraar, and Zweitering, Phgszca’c G a s . , 17, 880 (1961). (6) W.F. Giauque and J. 0. Clayton, J . Amer. Chem. SOC.,55, 4879 (1933). (7) C. J. Egan and J. D. Kemp, ibid., 59, 1264 (1937). (8) A. D.Crowell and D. M. Young, Trans. Faradau A’OC., 49, 1080 (1953). (9) A. J. Eatwell and B. L. Smith, Phil. Mag., 5, 463 (1960). (10) J. Lander and J. Morrison, Surface Sci., 5, 163 (1966).
971
ARGON-ETHYLENE LAYERFORMATION ON GRAPHITIZED CARBONBLACK
14
I2
p'
'I z < a
IO
t
u 0
8
0 W
m K
5 U
6
W
2
3
6 >
4
,2
0
0. I
0.2
0.3
0.4
0.6
0.5
0.7
0.8
0.9
I .o
P/fb Figure 1. Argon adsorption on various coverages of ethylene a t 77.5'K.
plane of graphite have so far not been published, we are unable to cross check the 11.6 m2/g value. Consequently, for the present we cautiously assign the true specific area of Sterling FT (2700) within the range 11.6 to 12.5 m2/g. Using a specific area of Sterling FT (2700) in this range and 2.44 cc/g as the monolayer volume of ethylene, we assign the area 17.8-18.8 A2 to each preadsorbed ethylene molecule. This range of area does not correspond with any of the areas calculated for localized adsorption at periodically spaced lattice sites on the
Table 11: Area per Ethylene Molecule in Various Orientations with Hexagonal Closest Packing Fixed orientation
Flat Vertical Sidewise Tilted at 35" Free rotating molecule
Area,
bz
19.2 12.3 15.2 16.4 20.6
graphitic carbon surface, such as the center of alternate hexagons of carbon atoms (20.8 A2). Furthermore, this range of area is significantly lower than the molecular mea of ethylene assigned by Kiselev" from BET measurements (21 A2) and even lower than the value he calculated for close packed ethylene molecules lying flat on the carbon black surface (19.5 A2). Table I1 shows the results of our molecular area calculations for closest packed ethylene in several possible fixed orientations with respect to the carbon black surface. I n addition to calculations for flat, verticaJ, and sidewise ethylene orientation, we have made calculations for a tilted ethylene molecule because potential energy calculations, which are summarized in Tables I11 and IV and which allow for the anisotropy of the ethylene molecule, indicate that the most stable orientation for an individual ethylene molecule occurs when the molecular plane is tilted at about 35" and the carbon(11) A. G. Beaus, V. P. Dreving, and A. V. Kislev, Russ. J. Phus. Chem., 38, 69 (1964). Volume YS,Number 4 April 1969
C. F. PRENZLOW, H. R. BEARD,AND R. S. BRUNDAGE
972 Table 111: Orientation os. Potential Energy of Adsorbed Ethylene Molecules
cx Orielitation Flat, molecular plane 1 I to surface; C=C bond 1 1 to surface Sidewise, molecular plane I, C=C bond 11 to surface Vertical, molecular plane I, C=C bond Ito surface Tilted, molecular plane 35" C=C bond 1 1 to surface
1068,
erg cmo
Z,
A
E , cal/mol
1.55
3.70
-2590
1.59
3.97
-2170
1.63
4.29
- 1760
1.68
3.55
-3180
Table IV: Parameters for Calculating Kirkwood-Mueller Constants
Ethylene Vertical orientation Sidewise orientation Flat orient at ion 35" tilt Random orientation Argon Graphite
52.5 39.0 33.8 35.8 42.6" 16.2 10.2
3.19 3.46 3.66 3.59 3.12b 3.24 13.5O
a J. Hirchfelder, C. Curtiss, and R. Bird, "Molecular Theory of Gases and Liquids," John Wiley & Sons, New York, N. Y., 1954, p 950. C. Barton, R. G. Meisenheimer, and D. P. Stevenson, J. Phys. Chern., 64, 1312 (1960). H. T . Pinkick Phgs. Rea., 94,319 (1954).
carbon bond is parallel to the carbon black surface. An end view of this Orientation is shown in Figure 2. These potential energy calculations were made using the equation
E -NnC/6Z3 (2) in which N is the number of carbon atoms per cc of carbon black, C is the Kirkwood-Mueller constant, and Z is the distance of closest approach of the ethylene molecule to the carbon black surface. The distance 2 was taken to be the distance from the center of the ethylene molecule in a particular orientation to the outermost plane of graphitic carbon black atoms. I n these calculations the C=C bond distance of ethylene was taken to be 1.33 A, the C-H distance 1.09 A, the =C